INDEX

Explanations

names of individuals and titles associated with renown or recognition

oai_token-act-pair · gpt-3.5-turbo

descriptors indicating a person's reputation or acclaim

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

adas

-0.83

agements

-0.81

NetMessage

-0.79

adra

-0.78

ossession

-0.77

blance

-0.75

raped

-0.73

ascript

-0.72

vertisement

-0.71

tiny

-0.70

POSITIVE LOGITS

 journalist

1.17

 writer

1.14

 scholar

1.13

 historian

1.12

 mathematician

1.11

 novelist

1.11

 exponent

1.10

 educator

1.10

 astroph

1.10

 author

1.08

Activations Density 0.168%

names of individuals and titles associated with renown or recognition

descriptors indicating a person's reputation or acclaim

No Comments

No Known Activations